Where is the API reference?

[havietisov]

I cannot find API reference anywhere, so how am I supposed to use it?

[LiuXiaoxuanPKU]

Quickstart might be helpful: https://vllm.readthedocs.io/en/latest/getting_started/quickstart.html
You can also checkout supported models here: https://vllm.readthedocs.io/en/latest/models/supported_models.html
The examples and benchmarks folders also include multiple use cases: https://github.com/vllm-project/vllm/tree/main/examples, https://github.com/vllm-project/vllm/tree/main/benchmarks

[havietisov]

Eh, well, I would like to see exactly the API reference. For instance, on how to use SamplingParams, what arguments are there, etc.

[kir152]

vllm servers are built on top of FastAPI. This means that servers created with FastAPI will automatically have the API documentation available at the /docs endpoint.
eg:
if you are running a vllm server locally, go to http://localhost:8080/docs

[hmellor]

https://docs.vllm.ai/en/latest/dev/sampling_params.html
https://docs.vllm.ai/en/latest/dev/engine/llm_engine.html
https://docs.vllm.ai/en/latest/dev/engine/async_llm_engine.html